Fig. 4: The brain LV from the behavioral-PLS analysis linking the psychopathology and physiology variates from CCA 1 (cf. Figure 3) to neural sexual differentiation in the ABCD sample. | Nature Communications

Fig. 4: The brain LV from the behavioral-PLS analysis linking the psychopathology and physiology variates from CCA 1 (cf. Figure 3) to neural sexual differentiation in the ABCD sample.

From: Genetic risk predicts adolescent mood pathology via sexual differentiation of brain function and physiological aging

Fig. 4

Panel A shows the correlations of the two CCA 1 variates with the Time 1 and Time 2 brain LV scores in the discovery PLS analysis. Panel B shows the correlations of the two CCA 1 variates with the predicted Time 1 and Time 2 brain LV scores (based on the 10-fold cross-validation procedure). A red asterisk indicates a robust correlation between the respective CCA 1 variate and the discovery (panel A) or predicted (panel B) brain LV scores across all participants. The violin plots in panels A and B depict the distribution of these correlation coefficients across the 100,000 bootstrap samples from the discovery (panel A) or cross-validated (panel B) PLS analysis. Panel C depicts the ROI-specific weights/loadings on the brain LV identified with the discovery PLS analysis with a bootstrap ratio greater than 2.75 in absolute value (equivalent to a 99% CI). Panel D depicts the Schaefer ROIs robustly correlated (based on cross-validated 99% confidence intervals) with the predicted value of the brain LV from the cross-validation procedure. These are partial correlations controlling for the confounders listed under “Control variables”. To facilitate interpretation, panels E and F present Schaefer network-based distributions of PLS weights (panel E) or partial correlations (panel F) summarizing the ROI-specific results from panels (C) and (D), respectively. The subnetworks from the Schaefer 17-network atlas (e.g., Control A/B/C) have been combined into one to increase comparability with the Gordon atlas. In panels C–F, positive values indicate masculinization, whereas negative values indicate feminization of the connectivity patterns associated with a specific ROI. As stated in the main text, sexual differentiation was estimated in reference to resting state connectivity data from the Human Connectome Project. BSR = bootstrap ratio. PLS = partial least squares. LV= latent variable. T1 Time 1, T2 Time 2, Schaefer networks: TP temporo-parietal, SAL-VAN salience/ventral attention, LB limbic, DMN default mode, DAN  dorsal attention, SM-A somatomotor-A, SM-B somatomotor-B. VIS visual.

Back to article page